Imp is a framework for typed, optimizable language-model programs on the BEAM. This release sends a chat message's participant name (the OpenAI chat format's optional name) to the provider, where Imp used to drop it, and lets a program name the messages of its history turns and its request. A message without a name is sent exactly as before, so no caller that does not name messages sees any change. The new input keys are public API a caller can rely on, so under Imp's versioning rule this is a minor release.

Install

{:imp, "~> 0.17.0"}

~> 0.17.0 takes later 0.17 patch releases and nothing newer. The two-part form, ~> 0.17, means >= 0.17.0 and < 1.0.0 in Mix, so mix deps.update imp under it takes every later 0.x release, including minor releases that change what callers receive.

Every dependency comes from Hex. Use a path dependency only while developing against a local checkout.

ExMCP and erlexec are declared runtime: false, so an OTP release that uses Imp.ACP or Imp.MCP must list applications: [ex_mcp: :load, erlexec: :load] in its release definition; see releases that use MCP or ACP. Ordinary Imp startup starts no protocol endpoint.

Changes and upgrading from 0.16.1

Change the Imp dependency to {:imp, "~> 0.17.0"}, run mix deps.update imp, and commit mix.lock. No code or data migration is needed. A program that already puts name on a message it gives Imp.LM, or :user_name in its inputs, now sends it.

  • Added: participant names reach the provider. Imp.Clients.ReqLLM copies a message's name (atom or string key) onto the message it sends, for generated, streamed and batched requests. Before, Imp.LM.generate(lm, [%{role: "user", content: "hi", name: "alice"}]) went out with no name. A provider whose format has no such field, such as Anthropic's Messages API, does not receive it.

  • Added: naming messages through the adapters. With Imp.Adapter.Chat, and the JSON, XML and PlanFirst adapters built on it, a history turn names its user message with :user_name and its assistant message with :assistant_name, beside its fields, and the request inputs name the final user message with :user_name:

    history = Imp.History.new([%{question: "Who won?", answer: "Blue.", user_name: "alice"}])
    Imp.call(predict, %{question: "Rematch?", history: history, user_name: "bob"})

    The keys are never rendered as fields, and tool results and notes are never named. A signature that declares a field called user_name or assistant_name keeps it as an ordinary field. Imp.Predict does not warn about :user_name as an extra input under those adapters. ReActV2 names the turns of a history it is given, but not its request message.

  • Not every model sees the name. Many upstreams accept name and drop it, among them Google's, DeepSeek's, Kimi's, Mistral's and Grok's, and open-weight models served through hosts such as DeepInfra. When the model must know who is speaking, say so in the content too.

Rollback means restoring {:imp, "~> 0.16.0"} and 0.16.1 in the lock; names are then dropped again and :user_name in the inputs warns as an extra input. See the changelog for every change.

Known limits

  • The usage map (:usage on the :model_response event, Imp.Usage, Imp.Prediction.get_lm_usage/1) is ReqLLM's, with input_includes_cached added. Its :cost and :total_cost are ReqLLM's catalog estimate, not a charge, for streamed and non-streamed calls alike, and OpenRouter's charge is its "cost". Read money from cost and estimated_cost.
  • An error the provider sends inside a stream has status nil and retryable: true, because ReqLLM's stream decoder keeps only its message; the same error in a non-streamed response may carry a status that says not to retry.
  • A flat two-element name-first list that is not itself an element of a list (at the top level, in a tuple, or under a key that is not a credential name), such as ["password", "hunter2"], is not redacted, where 0.5.0 redacted it. Such a list is read as a key and value only as an element of a list, so that lists of names such as with_inputs([:api_key, :question]) survive saving.
  • The GEPA, SIMBA, MIPROv2, InferRules and random-search checkpoints redact their failure reasons but otherwise hold the resume state as it is: GEPA's candidates, evaluation cache, proposed instructions and reflection data, and the others' instructions, demos and scores. Fast-Slow and Playbook checkpoints likewise hold their prompts, rollouts and playbooks. Treat checkpoint files as sensitive.
  • A ReActV2 step whose prose quotes its own field names (a JSON object with next_thought or tool_calls keys, a [[ ## field ## ]] line, a <next_thought> tag) is not read as prose and goes to the JSON fallback, which usually costs one more call. If the model repeats the same reply, the step ends :incomplete. Beside native tool calls, such a thought becomes nil; the calls still run.
  • Tool images reach the model as user messages after the turn's tool responses, labelled with the call, not nested in the tool result, even for providers that accept images in tool results. Only MCP text and image blocks are converted; other block types stay text, image URLs in text are never fetched, and the selected model must accept image input of that format and size.
  • ATIF's extra.seen_by_model and the read-as content come from matching a tool message in the next captured request. A result that request does not carry, including one that ended the run, keeps the tool's own output and is marked unseen. agent.tool_definitions comes only from :tools_sent events, and final_metrics totals are left out when any response lacks that figure.
  • A streamed completion that the provider's content filter stops still ends as a completion: the stream's finish_reason is :content_filter and the filter's text has already arrived as chunks. Only non-streamed requests turn it into an Imp.LMError.
  • When a streamed ReqLLM request fails, ReqLLM's own warning log prints the error with its response headers. Imp strips them from the error it returns, but cannot change that log line.
  • In command output, a Bearer token of 12 to 15 characters, or one with no digit, is shown when more text or a newline follows it. 0.5.0's command pipeline hid any such token of 12 or more characters.
  • A GEPA checkpoint taken between preparing and starting a full validation resumes by rejecting that candidate, as before.

The CHANGELOG records every user-visible change in this release. Generated module documentation is the complete API reference. Start with Imp, Imp.Signature, Imp.Module, Imp.Evaluate, Imp.Optimizer, Imp.ACP, Imp.MCP, and Imp.Telemetry.